Mutations in oncogenes and tumor suppressor genes in animal and human tumors may give important clues as to the exposures that led to the tumors. For rodent neoplasms that are pathologically similar to the corresponding human cancer, the rodent disease may be used to model the cellular events involved and to study prevention and therapy. Currently, we are focusing on the K-ras gene in lung cancer. In studies of the role of the oncogene K-ras in genesis of adenocarcinoma of the lung, we are utilizing a mouse model, including tumors induced in vivo, and normal immortalized and transformed alveolar type 2 cells in vitro. We have accumulated considerable evidence that K-ras is actually a tumor suppressor gene in the lung cells which can become adenocarcinoma. In lung tumors caused by N-nitrosodimethylamine and promoted by dioxin, amounts of total and membrane (active) K-ras protein were much reduced. Dioxin is an important environmental contaminant and human carcinogen for the respiratory tract. Thus, part of its mechanism of action may be downregulation of the tumor suppressive action of K-ras. This study also revealed for the first time that the proto-oncogene c-raf-1 is focally upregulated in developing tumors, and that raf-expressing macrophages become associated with pulmonary neoplasms from an early stage. These results lead to an important question: why is mutant K-ras aggressively oncogenic? Our results suggest that generation of reactive oxygen species is involved, leading to DNA damage, and that the enzyme COX2 may be part of the pathway. These findings may lead to new strategies for lung cancer prevention and therapy.